Over the past five years, a lot has changed, but one thing that’s remained constant since we last wrote about it in 2017 is that there’s really no such thing as a “tech industry.” The exact contours or label of a given industry may sound trivial. But the assumption that the major tech-driven companies operate substantially the same way, such that they belong to the same industry, is a potent one that could lead to some unfortunate results.

The announcement that Sens. Lindsey Graham (R-SC) and Elizabeth Warren (D-MA) are working together on a proposal to regulate social media platforms helps illustrate some of the dangers of treating all tech companies the same. In commentary during an oversight hearing examining allegations from Twitter’s former chief security official, Sen. Graham outlined the idea, which is aimed at curbing tech platform abuses. At first glance, the concept he described appears limited to a licensure regime for social media companies, and he didn’t specifically mention other kinds of tech companies. However, the idea’s cheerleaders are seeing the effort as an opportunity to enact a long sought-after goal: a single regulator for all large tech companies.

When we testified before the Senate’s Antitrust subcommittee in March 2020, our written comments distinguished between software platforms—like app stores—and other kinds of tech platforms, like social media, retail, and search. In their current form, the major social media platforms’ primary offerings are free to consumers and advertisers pay the bills. To maximize revenue, social media platforms have an incentive to keep users on their sites to push ad impressions upwards and increase the probability of converting an impression to a sale.

Meanwhile, the software platform revenue model differs markedly from social media’s insofar as consumers pay directly for smart devices and the products and services they access through the app stores on those devices. And the retail platform model differs from both software and social media platforms, though it arguably has more in common with software distribution. Now, think about a regulatory regime for social media companies—designed to address the inherent tension between ad revenue maximization and user welfare, not to mention all the other intractable content moderation issues—and try to apply it to software and retail distribution. Are app stores or retail platforms accused of funneling terabytes of inflammatory content to maximize consumer rage and sadness to extend the amount of time they spend on their platforms? Not exactly. So, why would we prescribe a remedy for app stores and retail platforms that’s designed for social media?

ACT | The App Association members are equally untethered to a “tech industry.” For example, Kaia Health is a digital therapeutic with a pain management focus, SwineTech is a pig farming management platform, and Quokka is a mobile security and privacy solutions company. The founders and employees of these companies are leveraging smart devices and software platforms to solve problems in a way that first and foremost requires specialized knowledge in healthcare, agriculture, and cybersecurity respectively. Putting all these companies under a single data privacy and security regime would be a sensible outcome, since it would be limited to specific kinds of activities: data collection, processing, and protection. Such an approach appropriately recognizes that technologies are tools and would establish guardrails around how those tools are used rather than dismantling or reshaping them completely. However, putting all of these companies under a single licensure scheme regulating their entire businesses would be a mistake, because Kaia’s responsibilities to patients and caregivers are vastly different from those of SwineTech to farmers. Thinking of all of these firms as belonging to a single set of “tech industry” companies puts the innovations they bring to the market at risk by applying inapposite remedies in unforeseen ways. Policymakers are better off abandoning “tech industry” assumptions.